Table 4. Associations of carotid measurements with genotype-predicted alcohol intake, and with ALDH2-rs671 and ADH1B-rs1229984 genotypes, in men and women.

	Effect per 280 g/week genotype-predicted mean male^a alcohol intake (95% CI)	P-value	P_het^b	ALDH2-rs671 GG vs AG Effect (95% CI)	P-value	P_het^b	ADH1B-rs1229984 GG vs AG Effect (95% CI)	P-value	P_het^b
cIMT, mm
Men	-0.008 (-0.018, 0.003)	0.148		-0.002 (-0.009, 0.004)	0.508		0.003 (-0.008, 0.013)	0.631
Women	-0.004 (-0.011, 0.002)	0.186	0.589	-0.004 (-0.008, 0.001)	0.090	0.738	0.004 (-0.003, 0.011)	0.267	0.855
Carotid plaque, OR
Men	1.21 (0.99, 1.49)	0.061		1.03 (0.92, 1.16)	0.573		1.06 (0.87, 1.28)	0.572
Women	0.98 (0.81, 1.17)	0.806	0.120	0.93 (0.84, 1.03)	0.160	0.178	0.97 (0.83, 1.15)	0.740	0.518
Carotid plaque burden, mm
Men	0.09 (0.02, 0.17)	0.018		0.05 (0.01, 0.10)	0.027		-0.02 (-0.10, 0.06)	0.591
Women	-0.01 (-0.06, 0.03)	0.600	0.022	-0.02 (-0.05, 0.01)	0.195	0.010	-0.03 (-0.08, 0.02)	0.247	0.883

Analyses were adjusted for age, and genomic principal components within each study area, and then combined by inverse-variance-weighted meta-analysis to yield the overall area-stratified genetic associations. The genetic instrument strength was assessed in 60,984 men with genotype information: Main genetic instrument F-statistic 1752 (range by area 43-783), variance in alcohol intake explained (r²) 13.6% (1.2%-22.5%); ALDH2-rs671 F-statistic 3267 (31–891), r² 10.5% (1.0%-22.6%); ADH1B-rs1229984 F-statistic 191 (4–43), r² 0.7% (0.1%-1.4%).

As women consumed little alcohol, the same six genotypic-area categories in women were used to estimate the genotypic effects in the same way as in men, in order to evaluate potential pleiotropic effects by comparing effects in men (who drank alcohol) with women.

P-value for heterogeneity between men and women.